MU — Micron Technology, Inc.

The market is bullish on a supercycle narrative, not a steady-eddie story. The 2026-09-30 FQ4 report blew past expectations ($54.23B revenue vs ~$51B consensus; non-GAAP EPS $33.42 vs $31.61), and FQ1 2027 guidance of $61.5B revenue / $38.15 EPS came in above consensus. Sell-side reaction was aggressive: DA Davidson raised its target to a Street-high $2,100 and Rosenblatt to $1,900, with the consensus 12-month target at $1,554.23 (12 analysts) and the recent-average target at $1,464.17 — 36%-45% above the 2026-10-04 price. Sentiment cooled only modestly (the stock slipped 2.05% on 2026-10-04 and -0.68% over the prior week) because of the sharp capex acceleration flagged at the print. Rating breakdown is 58 buy / 10 hold / 2 sell.

The trend is vertical, not merely accelerating: quarterly revenue ran $4.73B (period ended 2023-11-30) → $7.75B (2024-08-29) → $8.71B (2024-11-28) → $8.05B (2025-02-27) → $9.30B (2025-05-29) → $11.32B (2025-08-28) → $13.64B (2025-11-27) → $23.86B (2026-02-26) → $41.46B (2026-05-28) → $54.23B (2026-09-03). That is two consecutive ~75% sequential jumps and +379% YoY in the latest quarter. Margins went from negative to extraordinary: gross margin -0.7% in the 2023-11-30 quarter versus 86.8% in FQ4 2026 ($47.05B gross profit); operating margin -23.9% to 81.7%; net margin -26.1% to 69.5% ($37.70B net income); diluted EPS -$1.12 to $32.87. Free cash flow went from -$395M to $26.81B in the same quarter. One genuine blip worth isolating: revenue dipped sequentially from $8.71B to $8.05B in the 2025-02-27 quarter before the explosion — the rest is a clean accelerating up-cycle. The honest caveat is that this shape is indistinguishable, in real time, from a cycle peak: the deterministic score itself flags the +75% sequential move between 2025-11-27 and 2026-02-26 as potentially distorting trend comparisons.

As of 2026-10-04, MU traded at $1,074.89 on TTM EPS of $75.19, a 14.27x trailing P/E against a 5-year average P/E of 43.13x, with a PEG of 0.016 — optically very cheap. But the trailing numbers are peak-cycle. On FY2027 consensus EPS of $171.83 the forward P/E is only 6.25x, which reads less as a bargain and more as the market explicitly refusing to capitalize peak-cycle earnings. On sales the picture inverts: 9.12x P/S versus a 5-year average of 4.59x, and 8.78x P/B on $138.4B of equity. EV/EBITDA of 10.97x is the one multiple that looks genuinely reasonable rather than either euphoric or distressed. Net read: fairly valued to modestly cheap on earnings if the supercycle holds, and priced for perfection on sales/book if it does not — consistent with the deterministic score's 'Neutral vs. its own history' valuation context.

At $1,074.89 on 2026-10-04, the market is paying 6.25x FY2027 consensus EPS of $171.83 and 14.27x trailing EPS of $75.19. The critical asymmetry: FQ1 2027 guidance of $38.15 EPS annualizes to only ~$152.60, so the $171.83 consensus already embeds second-half acceleration that has not been delivered. That means you are underwriting another ~129% EPS increase on top of a year in which quarterly EPS already went from $2.83 to $32.87. If FY2027 lands at consensus and the stock re-rates to just 8x, that is ~$1,375 (28% above the 2026-10-04 price); at 7x it is ~$1,203 (12% above). If instead the cycle rolls and FY2028 EPS halves to ~$85, a 12x multiple gets you ~$1,020 — only ~5% below the current price, because the market has already discounted a great deal. The skew is positive, but it hinges entirely on the FQ1 print beating $38.15 and management guiding FQ2 higher again. This is a 'payoff is convex but conditional' setup, not a value trap and not a free lunch.

As of 2026-10-04, RSI(14) was 59.26 — neutral, and notably NOT overbought despite a six-month return of +193.5%. Price sat 4.87% above the 20-day SMA (implied ~$1,025), 12.43% above the 50-day SMA (implied ~$956) and 57.70% above the 200-day SMA (implied ~$682). The 52-week range is $179.61–$1,255, so the stock is near the top of an extraordinarily wide band — a good proxy for how violently this name moves in both directions. Short-term momentum has cooled (1-week -0.68%, day -2.05%) after post-earnings profit-taking. Entry-timing implication: the uptrend is intact but the extension versus the 200-day is extreme, so this is a poor spot to chase. A retest of the 20-day (~$1,025), and better still the 50-day (~$956), would be materially more attractive; a decisive break below the 50-day on heavy volume would be the first credible signal that the post-earnings move is unwinding.

• AI infrastructure capex cycle — demand is almost entirely a function of hyperscaler spending on HBM3e/HBM4 and data-center DRAM; MU is now a leveraged proxy on AI capex continuing to rise.
• Memory supply discipline — DRAM is effectively a three-player oligopoly and industry tightness has kept ASPs and margins elevated; any simultaneous capacity expansion would pressure pricing fast.
• Capital intensity arms race — quarterly capex jumped from $6.39B (period ended 2026-02-26) to $17.17B (2026-09-03) with further acceleration guided, the classic mechanism by which memory up-cycles end in oversupply.
• High beta / rate sensitivity — beta of 2.222 makes a $1.21T market cap trading at 9.1x sales acutely sensitive to rate repricing and risk-appetite swings; MU will underperform sharply in a hawkish rate shock.
• Geopolitics and export controls — China exposure, potential semiconductor tariffs and tightening restrictions on AI memory supply chains remain live policy risks given Micron's global fab footprint.

• FQ1 2027 results (next earnings 2026-12-16, estimated), guided to $61.5B revenue ±$1.5B and $38.15 adjusted EPS ±$1.00 — the guide already sits above consensus, so a beat-and-raise is required to sustain the momentum trade.
• Strategic Customer Agreement (SCA) execution — management expects SCAs to cover 35% of revenue through 2030; incremental signings or favorable price resets would structurally de-risk the cycle and support multiple expansion.
• Capital return announcement — FQ4 2026 alone produced $43.97B of operating cash flow and $26.81B of free cash flow, making a large buyback authorization or dividend increase a plausible near-term catalyst.
• HBM4 ramp and sold-out-capacity commentary through 2027 — confirmation of fully committed HBM supply extends the cycle narrative and underpins pricing.

• Cycle-peak mean reversion — 86.8% gross margin and 69.5% net margin in FQ4 2026 are unprecedented for a memory manufacturer (prior cycle peaks were roughly 58% in 2018 and 47% in 2021); reversion from here would be severe.
• Capex-driven oversupply — $17.17B of capex in a single quarter and still accelerating is the same setup that preceded every prior memory downturn; if all three DRAM makers expand at once, ASPs collapse.
• AI demand elasticity — 379% YoY revenue growth depends on hyperscaler AI build-outs continuing; an AI capex slowdown would hit volume and price simultaneously, and FY2027's $171.83 consensus EPS is built on that demand persisting.
• Technical and volatility risk — up 193.5% in six months, 57.7% above the 200-day SMA, beta of 2.222, and a 52-week range of $179.61–$1,255 all demonstrate how far this can fall in a de-rating.
• Thin valuation cushion on sales and book — 9.12x P/S (vs. a 4.59x five-year average) and 8.78x P/B leave very little support if earnings normalize toward mid-cycle levels.

Micron at $1,074.89 on 2026-10-04 is a high-reward, high-risk cyclical bet rather than a comfortable compounder. The fundamentals are genuinely extraordinary and cash-backed — $54.23B of quarterly revenue, $26.81B of quarterly free cash flow, 50.4% ROIC, 61.4% ROE, $38.4B of cash against only $4.7B of long-term debt — and at 6.25x forward EPS and 10.97x EV/EBITDA the stock is not obviously expensive. What it is not is safe: 86.8% gross margins have no historical precedent, capex is exploding quarter over quarter, and the shares sit 57.7% above their 200-day average. I would rather build a position on a pullback toward the 50-day (~$956) than chase here, with a pre-committed plan to trim if the 2026-12-16 FQ1 print misses the $38.15 guide or if capex guidance escalates again. Lean positive, but size for a stock that can halve.

A two-leg blend, both anchored to FMP data as of 2026-10-04. Leg 1 (50% weight): 8.0x FY2027 consensus EPS of $171.83 = $1,374.60 — a deliberate premium to the ~5-6x the market historically pays on peak memory earnings, justified by contracted SCA revenue covering 35% of sales through 2030, a three-player DRAM oligopoly and sold-out HBM, but still a steep discount to the broad market. Leg 2 (50% weight): 12x a normalized through-cycle EPS of ~$105 (blending FY2026 actual EPS of $74.22, FY2027E of $171.83, and two conservative post-peak years of roughly $90 and $30) = $1,260. The average of the legs is $1,317, rounded to $1,320 — about 23% above the 2026-10-04 price of $1,074.89 but roughly 15% below the $1,554.23 sell-side consensus, reflecting cycle-peak risk that consensus targets underweight.